EDA¶

Load and transform data.

In [1]:
import pandas as pd
In [2]:
df = pd.read_csv('data/itunes_data.csv')
df['Minutes'] = df['Milliseconds'] / (1000 * 60)
df['MB'] = df['Bytes'] / 1000000
df.drop(['Milliseconds', 'Bytes'], axis=1, inplace=True)

Boxplots and Boxenplots¶

In [3]:
import matplotlib.pyplot as plt
In [4]:
df['Minutes'].plot.box()
plt.show()
No description has been provided for this image
In [5]:
# save figure for book
f = plt.figure(figsize=(5.5, 5.5))  # this changes the size of the image -- more on this is chapter 5
f.patch.set_facecolor('w')  # sets background color behind axis labels
df['Minutes'].plot.box()
plt.tight_layout()  # auto-adjust margins
No description has been provided for this image

This should show up automatially in Jupyter notebooks, if not, try running the magic command %matplotlib or %matplotlib inline in a code cell.

In [6]:
df['Minutes'].plot.box()
Out[6]:
<Axes: >
No description has been provided for this image
In [7]:
import seaborn as sns
_ = sns.boxenplot(y=df['Minutes'])
No description has been provided for this image
In [8]:
# plot multiple columns at once
sns.boxenplot(data=df[['Minutes', 'MB']])
Out[8]:
<Axes: >
No description has been provided for this image
In [9]:
# save figure for book
f = plt.figure(figsize=(5.5, 5.5))  # this changes the size of the image -- more on this is chapter 5
f.patch.set_facecolor('w')  # sets background color behind axis labels
sns.boxenplot(y=df['Minutes'])
plt.tight_layout()  # auto-adjust margins
No description has been provided for this image

To hide the text output, send it to the special variable _.

In [10]:
sns.boxenplot(y=df['Minutes'])
plt.yscale('log')
No description has been provided for this image
In [11]:
# save figure for book
f = plt.figure(figsize=(5.5, 5.5))  # this changes the size of the image -- more on this is chapter 5
f.patch.set_facecolor('w')  # sets background color behind axis labels
sns.boxenplot(y=df['Minutes'])
plt.yscale('log')
plt.tight_layout()  # auto-adjust margins
No description has been provided for this image
In [12]:
df['Minutes'].plot.box()
plt.yscale('log')
No description has been provided for this image
In [13]:
# another way to use a log scale
df['Minutes'].plot.box(logy=True)
Out[13]:
<Axes: >
No description has been provided for this image
In [14]:
df['Minutes'].describe()
Out[14]:
count    3503.000000
mean        6.559987
std         8.916757
min         0.017850
25%         3.454683
50%         4.260567
75%         5.360750
max        88.115883
Name: Minutes, dtype: float64

Violin Plots¶

In [15]:
sns.histplot(x=df['Minutes'], kde=True)
Out[15]:
<Axes: xlabel='Minutes', ylabel='Count'>
No description has been provided for this image
In [16]:
# save figure for book
f = plt.figure(figsize=(5.5, 5.5))  # this changes the size of the image -- more on this is chapter 5
f.patch.set_facecolor('w')  # sets background color behind axis labels
sns.histplot(x=df['Minutes'], kde=True)
plt.tight_layout()  # auto-adjust margins
No description has been provided for this image
In [17]:
sns.violinplot(data=df, x='Minutes')
Out[17]:
<Axes: xlabel='Minutes'>
No description has been provided for this image
In [18]:
# save figure for book
f = plt.figure(figsize=(5.5, 5.5))  # this changes the size of the image -- more on this is chapter 5
f.patch.set_facecolor('w')  # sets background color behind axis labels
sns.violinplot(data=df, x='Minutes')
plt.tight_layout()  # auto-adjust margins
No description has been provided for this image
In [19]:
top_5_genres = df['Genre'].value_counts().index[:5]
top_5_data = data=df[df['Genre'].isin(top_5_genres)]
In [20]:
sns.violinplot(data=top_5_data, x='Minutes', y='Genre')
Out[20]:
<Axes: xlabel='Minutes', ylabel='Genre'>
No description has been provided for this image
In [21]:
# save figure for book
f = plt.figure(figsize=(5.5, 5.5))  # this changes the size of the image -- more on this is chapter 5
f.patch.set_facecolor('w')  # sets background color behind axis labels
sns.violinplot(data=top_5_data, x='Minutes', y='Genre')
plt.tight_layout()  # auto-adjust margins
No description has been provided for this image

Scatter plots¶

In [22]:
plt.scatter(df['Minutes'], df['MB'])
Out[22]:
<matplotlib.collections.PathCollection at 0x7efc5efa79d0>
No description has been provided for this image
In [23]:
# save figure for book
f = plt.figure(figsize=(5.5, 5.5))  # this changes the size of the image -- more on this is chapter 5
f.patch.set_facecolor('w')  # sets background color behind axis labels
plt.scatter(df['Minutes'], df['MB'])
plt.tight_layout()  # auto-adjust margins
No description has been provided for this image
In [24]:
plt.scatter(df['Minutes'], df['MB'], alpha=0.1)
Out[24]:
<matplotlib.collections.PathCollection at 0x7efc5eea2990>
No description has been provided for this image
In [25]:
plt.scatter(df['Minutes'], df['MB'])
plt.xlabel('Minutes')
plt.ylabel('MB')
Out[25]:
Text(0, 0.5, 'MB')
No description has been provided for this image
In [26]:
plt.scatter(df['Minutes'], df['MB'])
plt.xlabel('Minutes')
plt.ylabel('MB')
Out[26]:
Text(0, 0.5, 'MB')
No description has been provided for this image
In [27]:
# save figure for book
f = plt.figure(figsize=(5.5, 5.5))  # this changes the size of the image -- more on this is chapter 5
f.patch.set_facecolor('w')  # sets background color behind axis labels
sns.scatterplot(data=df, x='Minutes', y='MB')
plt.tight_layout()  # auto-adjust margins
No description has been provided for this image
In [28]:
sns.scatterplot(data=df, x='Minutes', y='MB')
Out[28]:
<Axes: xlabel='Minutes', ylabel='MB'>
No description has been provided for this image
In [29]:
sns.scatterplot(data=top_5_data, x='Minutes', y='MB', hue='Genre')
Out[29]:
<Axes: xlabel='Minutes', ylabel='MB'>
No description has been provided for this image
In [30]:
# save figure for book
f = plt.figure(figsize=(5.5, 5.5))  # this changes the size of the image -- more on this is chapter 5
f.patch.set_facecolor('w')  # sets background color behind axis labels
sns.scatterplot(data=top_5_data, x='Minutes', y='MB', hue='Genre')
plt.tight_layout()  # auto-adjust margins
No description has been provided for this image

Pairplots and Correlograms¶

In [31]:
sns.pairplot(data=df)
Out[31]:
<seaborn.axisgrid.PairGrid at 0x7efc61de1fd0>
No description has been provided for this image
In [32]:
# save figure for book
f = plt.figure(figsize=(5.5, 5.5))  # this changes the size of the image -- more on this is chapter 5
f.patch.set_facecolor('w')  # sets background color behind axis labels
sns.pairplot(data=df)
plt.tight_layout()  # auto-adjust margins

<Figure size 550x550 with 0 Axes>
No description has been provided for this image
In [35]:
import numpy as np


sns.heatmap(df.select_dtypes(include=np.number).corr(), annot=True)
Out[35]:
<Axes: >
No description has been provided for this image
In [38]:
# save figure for book
f = plt.figure(figsize=(5.5, 5.5))  # this changes the size of the image -- more on this is chapter 5
f.patch.set_facecolor('w')  # sets background color behind axis labels
sns.heatmap(df.select_dtypes(include=np.number).corr(), annot=True)
plt.tight_layout()  # auto-adjust margins
No description has been provided for this image
In [40]:
sns.heatmap(df.select_dtypes(include=np.number).corr(method='spearman'), annot=True)
Out[40]:
<Axes: >
No description has been provided for this image
In [41]:
# save figure for book
f = plt.figure(figsize=(5.5, 5.5))  # this changes the size of the image -- more on this is chapter 5
f.patch.set_facecolor('w')  # sets background color behind axis labels
sns.heatmap(df.select_dtypes(include=np.number).corr(method='spearman'), annot=True)
plt.tight_layout()  # auto-adjust margins
No description has been provided for this image

Missing value plot¶

In [45]:
import missingno as msno
msno.matrix(df)
Out[45]:
<Axes: >
No description has been provided for this image
In [46]:
# save figure for book
f = msno.matrix(df, figsize=(5.5, 5.5))
f.patch.set_facecolor('w')  # sets background color behind axis labels
f2 = f.get_figure()
No description has been provided for this image

EDA packages - pandas-profiling¶

Install pandas profiling with pip instead of conda - problems with conda version

Could Not Install (using pip), sorry¶

In [49]:
import pandas_profiling
pandas_profiling.__version__

---------------------------------------------------------------------------
ModuleNotFoundError                       Traceback (most recent call last)
Cell In[49], line 1
----> 1 import pandas_profiling
      2 pandas_profiling.__version__

ModuleNotFoundError: No module named 'pandas_profiling'
In [ ]:
from pandas_profiling import ProfileReport
In [ ]:
report = ProfileReport(df)
In [ ]:
report.to_widgets()

Summarize dataset:   0%|          | 0/21 [00:00<?, ?it/s]
Generate report structure:   0%|          | 0/1 [00:00<?, ?it/s]
Render widgets:   0%|          | 0/1 [00:00<?, ?it/s]
VBox(children=(Tab(children=(Tab(children=(GridBox(children=(VBox(children=(GridspecLayout(children=(HTML(valu…
In [50]:
# create our own histogram of the length of the tracks
df['Track'].str.len().plot.hist(bins=50)
Out[50]:
<Axes: ylabel='Frequency'>
No description has been provided for this image
In [51]:
import numpy as np
df_log = df.copy()
df_log['log(Minutes)'] = np.log(df_log['Minutes'])
df_log['log(MB)'] = np.log(df_log['MB'])
sns.jointplot(x="log(Minutes)", y="log(MB)", data=df_log, kind="hex")
Out[51]:
<seaborn.axisgrid.JointGrid at 0x7efc61de3b60>
No description has been provided for this image
In [52]:
# saving figure for book
sns.jointplot(x="log(Minutes)", y="log(MB)", data=df_log, kind="hex")
plt.tight_layout()  # auto-adjust margins
No description has been provided for this image
In [54]:
report.to_notebook_iframe()

---------------------------------------------------------------------------
NameError                                 Traceback (most recent call last)
Cell In[54], line 1
----> 1 report.to_notebook_iframe()

NameError: name 'report' is not defined

Other EDA packages: autoviz, sweetviz, d-tale¶

In [55]:
from autoviz.AutoViz_Class import AutoViz_Class

---------------------------------------------------------------------------
ModuleNotFoundError                       Traceback (most recent call last)
File ~/CS-620/.venv/lib64/python3.13/site-packages/xgboost/compat.py:105
    104 try:
--> 105     import pkg_resources
    106     pkg_resources.get_distribution('dask')

ModuleNotFoundError: No module named 'pkg_resources'

During handling of the above exception, another exception occurred:

NameError                                 Traceback (most recent call last)
Cell In[55], line 1
----> 1 from autoviz.AutoViz_Class import AutoViz_Class

File ~/CS-620/.venv/lib64/python3.13/site-packages/autoviz/__init__.py:3
      1 name = "autoviz"
      2 from .__version__ import __version__, __holo_version__
----> 3 from .AutoViz_Class import AutoViz_Class
      4 from .AutoViz_Class import data_cleaning_suggestions
      5 from .AutoViz_Class import FixDQ

File ~/CS-620/.venv/lib64/python3.13/site-packages/autoviz/AutoViz_Class.py:29
     27 from pandas_dq import Fix_DQ, dq_report
     28 ##########################################################################################
---> 29 from autoviz.AutoViz_Holo import AutoViz_Holo
     30 from autoviz.AutoViz_Utils import draw_pivot_tables, draw_scatters
     31 from autoviz.AutoViz_Utils import draw_pair_scatters, draw_barplots, draw_heatmap

File ~/CS-620/.venv/lib64/python3.13/site-packages/autoviz/AutoViz_Holo.py:3
      1 from holoviews.ipython import display
----> 3 from autoviz.AutoViz_Utils import classify_print_vars, find_remove_duplicates
      4 import numpy as np
      5 import pandas as pd

File ~/CS-620/.venv/lib64/python3.13/site-packages/autoviz/AutoViz_Utils.py:2098
   2092     return train, test
   2095 ###################################################################################
   2096 ################      Find top features using XGB     ###################
   2097 ################################################################################
-> 2098 from xgboost import XGBClassifier, XGBRegressor
   2101 def find_top_features_xgb(train, preds, numvars, target, modeltype, corr_limit=0.7, verbose=0):
   2102     """
   2103     This is a fast utility that uses XGB to find top features.
   2104     It returns a list of important features.
   2105     Since it is XGB, you don't have to restrict the input to just numeric vars.
   2106     You can send in all kinds of vars and it will take care of transforming it. Sweet!
   2107     """

File ~/CS-620/.venv/lib64/python3.13/site-packages/xgboost/__init__.py:6
      1 """XGBoost: eXtreme Gradient Boosting library.
      2 
      3 Contributors: https://github.com/dmlc/xgboost/blob/master/CONTRIBUTORS.md
      4 """
----> 6 from .core import (
      7     DMatrix,
      8     DeviceQuantileDMatrix,
      9     Booster,
     10     DataIter,
     11     build_info,
     12     _py_version,
     13 )
     14 from .training import train, cv
     15 from . import rabit  # noqa

File ~/CS-620/.venv/lib64/python3.13/site-packages/xgboost/core.py:21
     18 import numpy as np
     19 import scipy.sparse
---> 21 from .compat import STRING_TYPES, DataFrame, py_str, PANDAS_INSTALLED
     22 from .libpath import find_lib_path
     23 from ._typing import (
     24     CStrPptr,
     25     c_bst_ulong,
   (...)     36     CupyT,
     37 )

File ~/CS-620/.venv/lib64/python3.13/site-packages/xgboost/compat.py:108
    106     pkg_resources.get_distribution('dask')
    107     DASK_INSTALLED = True
--> 108 except pkg_resources.DistributionNotFound:
    109     dask = None
    110     DASK_INSTALLED = False

NameError: name 'pkg_resources' is not defined

ERROR: Could not find a version that satisfies the requirement pkg_resources (from versions: none) ERROR: No matching distribution found for pkg_resources

[notice] A new release of pip is available: 23.2.1 -> 25.1.1 [notice] To update, run: pip install --upgrade pip

In [56]:
# the documentation is not great for autoviz, but you can use a dataframe like so
AutoViz_Class().AutoViz(filename="", dfte=df)

---------------------------------------------------------------------------
NameError                                 Traceback (most recent call last)
Cell In[56], line 2
      1 # the documentation is not great for autoviz, but you can use a dataframe like so
----> 2 AutoViz_Class().AutoViz(filename="", dfte=df)

NameError: name 'AutoViz_Class' is not defined

Sweetviz¶

In [ ]:
In [57]:
import sweetviz as sv

sv = sv.analyze(df)
sv.show_notebook()

---------------------------------------------------------------------------
ModuleNotFoundError                       Traceback (most recent call last)
Cell In[57], line 1
----> 1 import sweetviz as sv
      3 sv = sv.analyze(df)
      4 sv.show_notebook()

File ~/CS-620/.venv/lib64/python3.13/site-packages/sweetviz/__init__.py:15
     12 __license__ = "MIT"
     14 # These are the main API functions
---> 15 from sweetviz.sv_public import analyze, compare, compare_intra
     16 from sweetviz.feature_config import FeatureConfig
     18 # This is the main report class; holds the report data
     19 # and is used to output the final report

File ~/CS-620/.venv/lib64/python3.13/site-packages/sweetviz/sv_public.py:4
      1 from typing import Union, List, Tuple
      2 import pandas as pd
----> 4 import sweetviz.dataframe_report
      5 from sweetviz.feature_config import FeatureConfig
      8 def analyze(source: Union[pd.DataFrame, Tuple[pd.DataFrame, str]],
      9             target_feat: str = None,
     10             feat_cfg: FeatureConfig = None,
     11             pairwise_analysis: str = 'auto'):

File ~/CS-620/.venv/lib64/python3.13/site-packages/sweetviz/dataframe_report.py:10
      8 from sweetviz.sv_types import NumWithPercent, FeatureToProcess, FeatureType
      9 import sweetviz.from_dython as associations
---> 10 import sweetviz.series_analyzer as sa
     11 import sweetviz.utils as su
     12 from sweetviz.graph_associations import GraphAssoc

File ~/CS-620/.venv/lib64/python3.13/site-packages/sweetviz/series_analyzer.py:4
      2 from sweetviz.sv_types import NumWithPercent, FeatureType, FeatureToProcess
      3 from sweetviz.type_detection import determine_feature_type
----> 4 import sweetviz.series_analyzer_numeric
      5 import sweetviz.series_analyzer_cat
      6 import sweetviz.series_analyzer_text

File ~/CS-620/.venv/lib64/python3.13/site-packages/sweetviz/series_analyzer_numeric.py:3
      1 import numpy as np
      2 import pandas as pd
----> 3 from sweetviz.graph_numeric import GraphNumeric
      4 import sweetviz.sv_html as sv_html
      5 from sweetviz.sv_types import NumWithPercent, FeatureType, FeatureToProcess

File ~/CS-620/.venv/lib64/python3.13/site-packages/sweetviz/graph_numeric.py:8
      5 import warnings
      7 from sweetviz.config import config
----> 8 from sweetviz import sv_html_formatters
      9 from sweetviz.sv_types import FeatureType, FeatureToProcess
     10 import sweetviz.graph

File ~/CS-620/.venv/lib64/python3.13/site-packages/sweetviz/sv_html_formatters.py:3
      1 from decimal import Decimal
      2 import numpy as np
----> 3 from sweetviz.graph_associations import CORRELATION_ERROR
      4 from sweetviz.graph_associations import CORRELATION_IDENTICAL
      7 def fmt_int_commas(value: float) -> str:

File ~/CS-620/.venv/lib64/python3.13/site-packages/sweetviz/graph_associations.py:6
      4 import matplotlib.pyplot as plt
      5 from sweetviz.sv_types import FeatureType
----> 6 import sweetviz.graph
      7 from sweetviz.config import config
      8 import itertools

File ~/CS-620/.venv/lib64/python3.13/site-packages/sweetviz/graph.py:8
      6 from io import BytesIO
      7 import base64
----> 8 from pkg_resources import resource_filename
      9 from pandas.plotting import register_matplotlib_converters
     11 from sweetviz import sv_html_formatters

ModuleNotFoundError: No module named 'pkg_resources'
In [58]:
import dtale
dtale.show(df)
Out[58]:
























Visualization best practices¶


















In [ ]:






df['Genre'].value_counts()[:5].plot.bar()
plt.xlabel('Genre')





















Out[ ]:




Text(0.5, 0, 'Genre')











No description has been provided for this image



















In [ ]:






df['Genre'].value_counts()[:5]





















Out[ ]:




Rock                  1297
Latin                  579
Metal                  374
Alternative & Punk     332
Jazz                   130
Name: Genre, dtype: int64




















In [ ]:






sns.countplot(y='Genre', data=df, order=df['Genre'].value_counts().index[:5], color='darkblue')





















Out[ ]:




<AxesSubplot:xlabel='count', ylabel='Genre'>











No description has been provided for this image



















In [ ]:






sns.countplot(y='Genre', data=df, order=df['Genre'].value_counts().index[:5])
plt.tight_layout()
























No description has been provided for this image























Choosing the right method for plotting¶


















In [ ]:






from sqlalchemy import create_engine
engine = create_engine('sqlite:///data/chinook.db')

with engine.connect() as connection:
    sql_df = pd.read_sql_table('invoices', connection)





















In [ ]:






sql_df.head()





















Out[ ]:












InvoiceId
CustomerId
InvoiceDate
BillingAddress
BillingCity
BillingState
BillingCountry
BillingPostalCode
Total






0
1
2
2009-01-01
Theodor-Heuss-Straße 34
Stuttgart
None
Germany
70174
1.98




1
2
4
2009-01-02
Ullevålsveien 14
Oslo
None
Norway
0171
3.96




2
3
8
2009-01-03
Grétrystraat 63
Brussels
None
Belgium
1000
5.94




3
4
14
2009-01-06
8210 111 ST NW
Edmonton
AB
Canada
T6G 2C7
8.91




4
5
23
2009-01-11
69 Salem Street
Boston
MA
USA
2113
13.86


























In [ ]:






sql_df.groupby('BillingCountry').sum().sort_values(by='Total', ascending=False)[:3]





















Out[ ]:












InvoiceId
CustomerId
Total




BillingCountry









USA
19103
2002
523.06




Canada
11963
1309
303.96




France
7168
1435
195.10


























In [ ]:






top_3_countries = sql_df.groupby('BillingCountry').sum().sort_values(by='Total', ascending=False)[:3].index.values





















In [ ]:






top_3_countries





















Out[ ]:




array(['USA', 'Canada', 'France'], dtype=object)




















In [ ]:






sql_df.set_index('InvoiceDate', inplace=True)





















In [ ]:






gb = sql_df[sql_df['BillingCountry'].isin(top_3_countries)]. \
        groupby([pd.Grouper(freq='M'), 'BillingCountry']).sum(). \
        groupby(level=-1).cumsum()





















In [ ]:






gb





















Out[ ]:













InvoiceId
CustomerId
Total




InvoiceDate
BillingCountry









2009-01-31
Canada
4
14
8.91




USA
5
23
13.86




2009-02-28
France
17
82
5.94




USA
18
39
14.85




2009-03-31
Canada
22
45
17.82




...
...
...
...
...




2013-10-31
USA
17477
1913
514.15




2013-11-30
France
7168
1435
195.10




USA
17882
1933
515.14




2013-12-31
Canada
11963
1309
303.96




USA
19103
2002
523.06






108 rows × 3 columns






















In [ ]:






gb.reset_index(inplace=True)





















In [ ]:






gb.head()





















Out[ ]:












InvoiceDate
BillingCountry
InvoiceId
CustomerId
Total






0
2009-01-31
Canada
4
14
8.91




1
2009-01-31
USA
5
23
13.86




2
2009-02-28
France
17
82
5.94




3
2009-02-28
USA
18
39
14.85




4
2009-03-31
Canada
22
45
17.82


























In [ ]:






sns.lineplot(data=gb, x='InvoiceDate', y='Total', hue='BillingCountry')





















Out[ ]:




<AxesSubplot:xlabel='InvoiceDate', ylabel='Total'>











No description has been provided for this image



















In [ ]:






# saving figure for book
sns.lineplot(data=gb, x='InvoiceDate', y='Total', hue='BillingCountry')
plt.tight_layout()
























No description has been provided for this image























Making plots redundant¶


















In [ ]:






# black and white redundancy
sns.lineplot(data=gb, x='InvoiceDate', y='Total', hue='BillingCountry', style='BillingCountry')





















Out[ ]:




<AxesSubplot:xlabel='InvoiceDate', ylabel='Total'>











No description has been provided for this image



















In [ ]:






# save figure for book
sns.lineplot(data=gb, x='InvoiceDate', y='Total', hue='BillingCountry', style='BillingCountry')
plt.tight_layout()
























No description has been provided for this image



















In [ ]:






sns.lineplot(data=gb, x='InvoiceDate', y='Total', hue='BillingCountry', style='BillingCountry', dashes=[(2, 1), (5, 2), ''])





















Out[ ]:




<AxesSubplot:xlabel='InvoiceDate', ylabel='Total'>











No description has been provided for this image



















In [ ]:






from labellines import labelLines





















In [ ]:






f = plt.figure()
ax = f.gca()
for country in top_3_countries:
    c_df = gb[gb['BillingCountry'] == country]
    ax.plot(c_df['InvoiceDate'], c_df['Total'], label=country)

labelLines(ax.get_lines())

plt.xlabel('Year')
plt.ylabel('Cumulative Sales ($)')

plt.tight_layout()
























No description has been provided for this image



















In [ ]:






sns.lineplot(data=gb, x='InvoiceDate', y='Total', hue='BillingCountry', size='BillingCountry', sizes=[1, 5, 1])





















Out[ ]:




<AxesSubplot:xlabel='InvoiceDate', ylabel='Total'>











No description has been provided for this image























Another way to collect and plot the same data¶


















In [ ]:






top_3_df_gb = sql_df[sql_df['BillingCountry'].isin(top_3_countries)].groupby([pd.Grouper(freq='M'), 'BillingCountry']).sum()
dfs = []
for country in top_3_countries:
    c_df = top_3_df_gb.xs(country, level=1).cumsum()
    c_df['Country'] = country
    dfs.append(c_df)
    
full_c_df = pd.concat(dfs)





















In [ ]:






full_c_df.head()





















Out[ ]:












InvoiceId
CustomerId
Total
Country




InvoiceDate










2009-01-31
5
23
13.86
USA




2009-02-28
18
39
14.85
USA




2009-03-31
80
121
28.71
USA




2009-04-30
106
140
42.57
USA




2009-06-30
220
205
61.38
USA


























In [ ]:






full_c_df.reset_index(inplace=True)





















In [ ]:






sns.lineplot(data=full_c_df, x='InvoiceDate', y='Total', hue='Country')





















Out[ ]:




<AxesSubplot:xlabel='InvoiceDate', ylabel='Total'>











No description has been provided for this image























Here is a start towards how we could collect the data using a SQL query.¶
We would need to either modify the SQL query to join the same data for the top 3 countries, or do 3 separate SQL queries and join the resulting dataframes.


















In [ ]:






engine = create_engine('sqlite:///data/chinook.db')

query = """SELECT SUM(Total) OVER (ORDER BY InvoiceId) as Total, strftime("%m-%Y", InvoiceDate) as 'month-year', BillingCountry
FROM invoices
WHERE BillingCountry="USA"
GROUP BY strftime("%m-%Y", InvoiceDate);
"""

with engine.connect() as connection:
    sql_df2 = pd.read_sql_query(query, connection)





















In [ ]:






sql_df2.head()





















Out[ ]:












Total
month-year
BillingCountry






0
13.86
01-2009
USA




1
14.85
02-2009
USA




2
16.83
03-2009
USA




3
30.69
04-2009
USA




4
34.65
06-2009
USA


























In [ ]:






# convert to datetime for better plotting
sql_df2['month-year'] = pd.to_datetime(sql_df2['month-year'])





















In [ ]:






sns.lineplot(data=sql_df2, x='month-year', y='Total', hue='BillingCountry')
plt.xlabel('Year')





















Out[ ]:




Text(0.5, 0, 'Year')











No description has been provided for this image























Saving Images¶


















In [ ]:






f = plt.figure(figsize=(5, 5))
sns.lineplot(data=gb, x='InvoiceDate', y='Total', hue='BillingCountry')
plt.tight_layout()
plt.savefig('cumulative_sales_lineplot.png', facecolor='w', dpi=300)
























No description has been provided for this image























Plotly¶


















In [ ]:






import plotly.express as px
px.line(gb, x='InvoiceDate', y='Total', color='BillingCountry', template='simple_white')




































 
  




















In [ ]:






# recall our original DataFrame has information on the tracks in our iTunes data
df.head()





















Out[ ]:












Track
Composer
UnitPrice
Genre
Album
Artist
Minutes
MB






0
For Those About To Rock (We Salute You)
Angus Young, Malcolm Young, Brian Johnson
0.99
Rock
For Those About To Rock We Salute You
AC/DC
5.728650
11.170334




1
Put The Finger On You
Angus Young, Malcolm Young, Brian Johnson
0.99
Rock
For Those About To Rock We Salute You
AC/DC
3.427700
6.713451




2
Let's Get It Up
Angus Young, Malcolm Young, Brian Johnson
0.99
Rock
For Those About To Rock We Salute You
AC/DC
3.898767
7.636561




3
Inject The Venom
Angus Young, Malcolm Young, Brian Johnson
0.99
Rock
For Those About To Rock We Salute You
AC/DC
3.513900
6.852860




4
Snowballed
Angus Young, Malcolm Young, Brian Johnson
0.99
Rock
For Those About To Rock We Salute You
AC/DC
3.385033
6.599424


























In [ ]:






px.scatter(df, x='Minutes', y='MB')

























 
  




















In [ ]:






px.histogram(df, x='Minutes')

























 
  




















In [ ]:






# recall this is our invoices table from the chinook iTunes database
sql_df.head()





















Out[ ]:












InvoiceId
CustomerId
BillingAddress
BillingCity
BillingState
BillingCountry
BillingPostalCode
Total




InvoiceDate














2009-01-01
1
2
Theodor-Heuss-Straße 34
Stuttgart
None
Germany
70174
1.98




2009-01-02
2
4
Ullevålsveien 14
Oslo
None
Norway
0171
3.96




2009-01-03
3
8
Grétrystraat 63
Brussels
None
Belgium
1000
5.94




2009-01-06
4
14
8210 111 ST NW
Edmonton
AB
Canada
T6G 2C7
8.91




2009-01-11
5
23
69 Salem Street
Boston
MA
USA
2113
13.86


























In [ ]:






gb_countries = sql_df.groupby('BillingCountry').sum()
gb_countries.reset_index(inplace=True)
gb_countries.head()





















Out[ ]:












BillingCountry
InvoiceId
CustomerId
Total






0
Argentina
1729
392
37.62




1
Australia
1043
385
37.62




2
Austria
1568
49
42.62




3
Belgium
1428
56
37.62




4
Brazil
7399
329
190.10


























In [ ]:






px.choropleth(gb_countries, locations="BillingCountry",
              locationmode='country names',
              color="Total")